High-throughput 3D random access three-photon structural and functional imaging Two-photon microscopy allows in vivo observation of neuronal dynamics at high spatial and temporal resolutions. The latest development of calcium indicators enables single action potential sensitivity and single dendritic spine resolution. However, such resolution and sensitivity is limited to the upper ~400 m of the neocortex in adult mice. Most of the studies are still restricted within layer 2/3 neurons. New methods are urgently in demand to investigate layer 5 and 6 neurons and subcortical structures. Recently, three-photon excitation has been employed to improve the imaging depth. With the longer excitation wavelength, the optical aberration and scattering are both reduced. Deep imaging at 1 mm depth has been demonstrated. To have efficient three-photon excitation, the laser pulse needs to have high peak intensity with the pulse energy at the focal plane above 1 nJ. Currently, optical parametric amplifier (OPA) with hundreds of kHz repetition rate is employed for three-photon excitation. A major drawback of the current implementation of three-photon microscopy is the low throughput, which is one order of magnitude less than that of conventional two-photon microscopy. In addition, the imaging has been constrained to a single 2D plane. To overcome the limitations of current three-photon microscopy, we propose to develop a high- throughput 3D random-access three-photon imaging system. First, our design will enable 3D imaging. As the neuronal population is distributed in 3D, the capability of 3D volume imaging will allow the coverage of significantly more neurons. Second, the random-access scanning can utilize the laser excitation and the measurement time efficiently. We will be able to achieve 10-20 times greater imaging throughput, which will be comparable or even higher than that of conventional resonant Galvo based two-photon imaging. Third, a major benefit of our new imaging system is the greatly improved imaging depth. From our in vivo measurement of mouse brain, our design can reach 1.7 mm with 1300 nm excitation, ~50% deeper than that of current three-photon microscopy systems. The high-throughput 3D random-access three-photon imaging system will be developed collaboratively by the engineers at Purdue and the neurobiologists at NYU. The neurobiologists will advise on the system design so that it is compatible with the routine neuroscience studies. Once completed, we will employ the system for calcium and structural imaging of subcortical brain regions. The ultimate goal is to have a turn-key solution that can be easily adopted by neurobiologists. We will make the system design (optics, optomechanics, data acquisition system, control software) freely available to the neuroscience community to quickly disseminate these methods.